一种基于经验模态分解和局部极大值点数的配电网高阻接地故障检测方法OA

The high impedance-grounding fault(HIF)of the distribution network is difficult to detect with conventional equipment due to its weak fault characteristics,which may cause safety issues such as overvoltage,fire,and personal electric shock.This paper introduces a HIF recognition method that combines time-frequency decomposition and threshold judgment.Initially,the zero-sequence current is processed using empirical mode decomposition(EMD).Subsequently,the second intrinsic mode function(IMF)is selected as the feature component and undergoes Gaussian filtering.Ultimately,the number of local maximum points of the feature component is counted to determine whether HIF has occurred.A 10kV distribution network model is constructed using PSCAD/EMTDC software to verify the algorithm.Experimental results indicate that the proposed method can effectively distinguish HIF from various common interference events with a high level of accuracy.Furthermore,it exhibits good adaptability under conditions of noise interference,changes in sampling rate,alterations in the grounding mode of the system's neutral point,and modifications to the system's structure.